...more recent posts
Still trying to get my ideas together regarding the communications device I have been thinking about. The main thing is that it is portable, wireless, and can work in two different modes. One mode is the standard internet mode, where the device searches for a gateway within range and if one is found it can connect through it to the internet. (I'm ignoring the cost structures, although there would be some costs here for someone.) The other mode would be a peer to peer mode. Each device would be a client and a server. If you are in peer mode then your device searches for any other peers within range. You would see some sort of list of available servers from which you could choose. As you physically move around the list would change as different people would come in or go out of range. On your own device you can either have your server on or off. If it's on, then people within range can connect to you. This would be like having your own apache (or whatever) server on an intranet. Also, in peer mode, your device is a router whether you have your server on or not. So people can connect through you to others. But there would be some (small) limit on the number of hops to avoid gnutella like problems with scaling. The peer mode would be a very local thing.
Apparently the different flavors of 802.11 (like Apple's Airport, or Lucent's Wavelan) can be used for things like this, so there is really nothing new here. I guess it just boils down to realizing that connecting to the global internet is not always what you want. This is what I realized in the hours following the WTC collapse, when a lot of Manhattan's communication infrastructure was knocked out. I wanted to talk to people outside the city, but contacting people who were here seemed even more pressing. Obviously if the attack had been more nebulous (like a bio attack) communicating locally would have been even more important. Being able to switch to peer mode, thus bypassing any fixed infrastructure (because the portable devices themselves are the infrastructure) would have been very convenient.
The internet itself is decentralized, but only somewhat. From what I understand about routing, this is an efficient design. Complete decentralization seems to have problems scaling because each node gets flooded with too many routing requests. I can only imagine that the internet itself is going to become much more regulated and business like in the near future. But as this happens perhaps a new frontier will open up in the space usually referred to as "the last mile." This space could be filled with an ad hoc peer to peer local intranet. This would be a community space - probably interesting in NYC and other cities, while less interesting or even non existent in very rural locations. The peer network would be free (in both senses) while the internet itself does and will continue to cost money. I'm actually in favor of a charge per byte structure on the global internet. Maybe being a gateway between local peer nets and the internet could be a business. People would pay to use your gateway to jump onto the net if they happened to be inside your peer network. And conversely, people could drop into local peer nets from remote locations using these gateways in the opposite direction.
In times of crisis the peer networks would work even if local infrastructure was knocked out. And then the government could keep a stash of disposable wireless gateways (something like 802.11 basestations) and maybe just drop them (on tiny parachutes or something) over a disaster area. Like seeding a field from the air. This way the local net would work right away, and connections to the global net could be gotten up and working very easily.
Or something like that. It's boring to read about it this way. I'm actually working on a short near future sci-fi story to flesh out this idea. Not sure if that will work out, but I'll put it up as I go if I make any progress.
Whenever I get back to the BCCI affair I know it's time to stop reading and go outside for a walk. There is no reason for me to think I can put the big picture together, so when these large pieces insist on falling into place I get a little nervous. Maybe this is a good time to dive into some difficult (for me) coding. Something with a lot of strands I can put together and then actuallly test as to whether I am right or not. Something that will either work or break. Maybe that mail interface, or the downloadable weblog snapshot feature. It's not as important as 60 briefcases that may or may not be missing, but at least I can work on it and then at some point reach a conclusion. I don't wish to learn any more about the Caspian Basin right now.